I'm having issues trying to form code for a problem I want to resolve. It goes like this:
~ Goal: flatten a nested list into one number
If the object is a list, replace the list with the sum of its atoms.
With nested lists, flatten the innermost lists first and work from there.
Example:
(CONDENSE '(2 3 4 (3 1 1 1) (2 3 (1 2)) 5))
(2 3 4 (6) (2 3 (3)) 5)
(2 3 4 (6) (8) 5)
(28)
=> 28
I've tried to implement the flatten list function for this problem and I ended up with this:
(defun condense (lst)
(cond
((null lst) nil)
((atom lst) (list lst)))
(t (append (flatten (apply #'+ (cdr lst))))))
But it gives me errors :(
Could anyone explain to me what is wrong with my processing/code? How can I improve it?
UPDATE: JUNE 5 2012
(defun condense(lxt)
(typecase lxt
(number (abs lxt))
(list
(if (all-atoms lxt)
(calculate lxt)
(condense (mapcar #'condense lxt))))))
So here, in this code, my true intent is shown. I have a function calculate that performs a calculation based off the values in the list. It is not necessarily the same operation each time. Also, I am aware that I am returning the absolute value of the number; I did this because I couldn't find another way to return the number itself. I need to find a way to return the number if the lxt is a number. And I had it recurse two times at the bottom, because this is one way that it loops on itself infinitely until it computes a single number. NOTE: this function doesn't implement a flatten function anymore nor does it use anything from it.
Imagine you have your function already. What does it get? What must it produce?
Given an atom, what does it return? Given a simple list of atoms, what should it return?
(defun condense (x)
(typecase x
(number
; then what?
(condense-number x))
(list
; then what?
(if (all-atoms x)
(condense-list-of-atoms x) ; how to do that?
(process-further-somehow
(condense-lists-inside x))))
; what other clauses, if any, must be here?
))
What must condense-lists-inside do? According to your description, it is to condense the nested lists inside - each into a number, and leave the atoms intact. So it will leave a list of numbers. To process that further somehow, we already "have" a function, condense-list-of-atoms, right?
Now, how to implement condense-lists-inside? That's easy,
(defun condense-lists-inside (xs)
(mapcar #'dowhat xs))
Do what? Why, condense, of course! Remember, we imagine we have it already. As long as it gets what it's meant to get, it shall produce what it is designed to produce. Namely, given an atom or a list (with possibly nested lists inside), it will produce a number.
So now, fill in the blanks, and simplify. In particular, see whether you really need the all-atoms check.
edit: actually, using typecase was an unfortunate choice, as it treats NIL as LIST. We need to treat NIL differently, to return a "zero value" instead. So it's better to use the usual (cond ((null x) ...) ((numberp x) ...) ((listp x) ...) ... ) construct.
About your new code: you've erred: to process the list of atoms returned after (mapcar #'condense x), we have a function calculate that does that, no need to go so far back as to condense itself. When you substitute calculate there, it will become evident that the check for all-atoms is not needed at all; it was only a pedagogical device, to ease the development of the code. :) It is OK to make superfluous choices when we develop, if we then simplify them away, after we've achieved the goal of correctness!
But, removing the all-atoms check will break your requirement #2. The calculation will then proceed as follows
(CONDENSE '(2 3 4 (3 1 1 1) (2 3 (1 2)) 5))
==
(calculate (mapcar #'condense '(2 3 4 (3 1 1 1) (2 3 (1 2)) 5)))
==
(calculate (list 2 3 4 (condense '(3 1 1 1)) (condense '(2 3 (1 2))) 5))
==
(calculate (list 2 3 4 (calculate '(3 1 1 1))
(calculate (list 2 3 (calculate '(1 2)))) 5))
==
(calculate (list 2 3 4 6 (calculate '(2 3 3)) 5))
==
(calculate (list 2 3 4 6 8 5))
==
28
I.e. it'll proceed in left-to-right fashion instead of the from the deepest-nested level out. Imagining the nested list as a tree (which it is), this would "munch" on the tree from its deepest left corner up and to the right; the code with all-atoms check would proceed strictly by the levels up.
So the final simplified code is:
(defun condense (x)
(if (listp x)
(reduce #'+ (mapcar #'condense x))
(abs x)))
a remark: Looking at that last illustration of reduction sequence, a clear picture emerges - of replacing each node in the argument tree with a calculate application. That is a clear case of folding, just such that is done over a tree instead of a plain list, as reduce is.
This can be directly coded with what's known as "car-cdr recursion", replacing each cons cell with an application of a combining function f on two results of recursive calls into car and cdr components of the cell:
(defun condense (x) (reduce-tree x #'+ 0))
(defun reduce-tree (x f z)
(labels ((g (x)
(cond
((consp x) (funcall f (g (car x)) (g (cdr x))))
((numberp x) x)
((null x) z)
(T (error "not a number")))))
(g x)))
As you can see this version is highly recursive, which is not that good.
Is this homework? If so, please mark it as such. Some hints:
are you sure the 'condensation' of the empty list in nil? (maybe you should return a number?)
are you sure the condensation of one element is a list? (maybe you should return a number?)
are you sure the condensation of the last case is a list? (shouldn't you return a number)?
In short, how is your condense ever going to return 28 if all your returned values are lists?
Task: With nested lists, flatten the innermost lists first and work from there
sum
flatten lists
For sum use REDUCE, not APPLY.
For flatten lists you need a loop. Lisp already provides specialized mapping functions.
Slightly more advanced: both the sum and the flatten can be done by a call to REDUCE.
You can also write down the recursion without using a higher-order function like APPLY, REDUCE, ... That's a bit more work.
Here's added the explanation of the errors you were having, actually you were close to solving your problem, just a bit more effort and you would get it right.
; compiling (DEFUN CONDENSE ...)
; file: /tmp/file8dCll3
; in: DEFUN CONDENSE
; (T (APPEND (FLATTEN (APPLY #'+ (CDR LST)))))
;
; caught WARNING:
; The function T is undefined, and its name is reserved
; by ANSI CL so that even
; if it were defined later, the code doing so would not be portable.
;
; compilation unit finished
; Undefined function:
; T
; caught 1 WARNING condition
;STYLE-WARNING: redefining CONDENSE in DEFUN
(defun condense (lst)
(cond
((null lst) nil)
((atom lst) (list lst)))
;.------- this is a function call, not a condition
;| (you closed the parens too early)
(t (append (flatten (apply #'+ (cdr lst))))))
;; Argument Y is not a NUMBER: (3 1 1 1)
;; [Condition of type SIMPLE-TYPE-ERROR]
(defun condense (lst)
(cond
((null lst) nil)
((atom lst) (list lst)); .-- not a number!
;You are calling #'+ -------. |
;on something, which | '(3 4 (3 1 1 1) (2 3 (1 2)) 5)
; is not a number. | |
(t (append (flatten (apply #'+ (cdr lst)))))))
;; You probably wanted to flatten first, and then sum
(defun condense (lst)
(cond
((null lst) nil); .--- returns just the
((atom lst) (list lst)); / atom 28, you can
; .---------------------/ just remove it.
(t (append (apply #'+ (flatten lst))))))
;; Now, you are lucky that append would just return the
;; atom if it's not a list
(defun condense (lst)
(cond
((null lst) nil)
((atom lst) (list lst))
(t (apply #'+ (flatten lst)))))
;; Again, you are lucky because (apply can take enough arguments
;; while your list is reasonably small - this will not always be
;; the case, that is why you need to use something more durable,
;; for example, reduce.
(defun condense (lst)
(cond
((null lst) nil)
((atom lst) (list lst))
(t (reduce #'+ (flatten lst)))))
;; Whoa!
(condense '(2 3 4 (3 1 1 1) (2 3 (1 2)) 5))
This is all given the flatten function actually works.
If your lisp already implements flatten and reduce functions (such as Clojure, which I will use here), you can just do something like:
user=> (defn condense [l] (reduce + 0 (flatten l)))
#'user/condense
user=> (condense [1 [2 [[3 4] 5]]])
15
user=>
Failing that, a naive implementation of those functions might be:
(defn flatten [l]
(cond (nil? l) l
(coll? l) (let [[h & t] l]
(concat (flatten h) (flatten t)))
true [l]))
and:
(defn reduce [op initial-value [h & t]]
(if (nil? t)
(op initial-value h)
(op initial-value (reduce op h t))))
But make sure to check the semantics of the particular Lisp you are using. Also, if you are implementing reduce and flatten, you may want to make them tail recursive which I didn't so as to maintain clarity.
In Common Lisp you would do something like:
(defun flatten (l)
(cond ((null l) l)
((atom l) (list l))
(t (append (flatten (car l))
(flatten (cdr l))))))
and use apply instead of reduce:
(defun condense (l) (apply #'+ (flatten l)))
Related
I just started to learn Common Lisp and this is my first functional programming language.
I am trying to learn about iterating through lists. I wrote these two functions:
(defun reverseList (liste)
(defvar reversedList(list))
(loop for i downfrom (-(length liste)1) to 0 do
(setf reversedList (append reversedList (list(nth i liste)))))
reversedList ;return
)
(defun countAppearance(liste element)
(defvar count 0)
(loop for i from 0 to (-(length liste) 1)do
(if (= (nth i liste) element)
(setf count (+ count 1))))
count
)
Both work fine for a regular list(ex: (1 3 5 7 3 9) but I want them to work for nested lists too.
Examples:
countAppearance - Input: (1 (3 5) (3 7 8) 2) 3 -> Expected output:2
reverseList - Input: (1 (2 3)) -> Expected output: ((3 2) 1)
Before I will show you solutions for nested lists, some notes about your code:
There is already function reverse for non-nested lists, so you don't have to reinvent the wheel.
=> (reverse (list 1 2 3 4 5))
(5 4 3 2 1)
If you need some local variables, use let or let*.
Lisp uses kebab-case, not camelCase, so rename reverseList as reverse-list and so on.
For (setf ... (+ ... 1)), use incf.
For iterating over list, use dolist.
Function count-occurrences can be written using recursion:
(defun count-occurrences (lst elem)
(cond ((null lst) 0)
((= (car lst) elem) (+ 1 (count-occurrences (cdr lst) elem)))
(t (count-occurrences (cdr lst) elem))))
CL-USER 3 > (count-occurrences (list 1 2 3 1 2 3) 2)
2
Or it can be written with let, dolist and incf:
(defun count-occurrences2 (lst elem)
(let ((count 0))
(dolist (e lst)
(when (= e elem) (incf count)))
count))
CL-USER 4 > (count-occurrences2 (list 1 2 3 1 2 3) 2)
2
Solutions for nested lists use recursion:
(defun deep-reverse (o)
(if (listp o)
(reverse (mapcar #'deep-reverse o))
o))
CL-USER 11 > (deep-reverse '(1 (2 3)))
((3 2) 1)
(defun deep-count (lst elem)
(cond ((null lst) 0)
((listp (car lst)) (+ (deep-count (car lst) elem)
(deep-count (cdr lst) elem)))
((= (car lst) elem) (+ 1 (deep-count (cdr lst) elem)))
(t (deep-count (cdr lst) elem))))
CL-USER 12 > (deep-count '(1 (3 5) (3 7 8) 2) 3)
2
Welcome to functional programming.
Firstly, there are some problems with the code that you have provided for us. There are some spaces missing from the code. Spaces are important because they separate one thing from another. The code (xy) is not the same as (x y).
Secondly, there is an important difference between local and global variables. So, in both cases, you want a local variable for reversedList and count. This is the tricky point. Common Lisp doesn't have global or local variables, it has dynamic and lexical variables, which aren't quite the same. For these purposes, we can use lexical variables, introduced with let. The keyword let is used for local variables in many functional languages. Also, defvar may not do what you expect, since it is way of writing a value once, which cannot be overwritten - I suspect that defparameter is what you meant.
Thirdly, looking at the reverse function, loop has its own way of gathering results into a list called collect. This would be a cleaner solution.
(defun my-reverse (lst)
(loop for x from (1- (length lst)) downto 0 collect (nth x lst)))
It can also be done in a tail recursive way.
(defun my-reverse-tail (lst &optional (result '()))
(if lst
(my-reverse-tail (rest lst) (cons (first lst) result))
result))
To get it to work with nested lists, before you collect or cons each value, you need to check if it is a list, using listp. If it is not a list, just add it onto the result. If it is a list, add on instead a call to your reverse function on the item.
Loop also has functionality to count items.
I'm trying to reverse a list in Lisp, but I get the error: " Error: Exception C0000005 [flags 0] at 20303FF3
{Offset 25 inside #}
eax 108 ebx 200925CA ecx 200 edx 2EFDD4D
esp 2EFDCC8 ebp 2EFDCE0 esi 628 edi 628 "
My code is as follows:
(defun rev (l)
(cond
((null l) '())
(T (append (rev (cdr l)) (list (car l))))))
Can anyone tell me what am I doing wrong? Thanks in advance!
Your code as written is logically correct and produces the result that you'd want it to:
CL-USER> (defun rev (l)
(cond
((null l) '())
(T (append (rev (cdr l)) (list (car l))))))
REV
CL-USER> (rev '(1 2 3 4))
(4 3 2 1)
CL-USER> (rev '())
NIL
CL-USER> (rev '(1 2))
(2 1)
That said, there are some issues with it in terms of performance. The append function produces a copy of all but its final argument. E.g., when you do (append '(1 2) '(a b) '(3 4)), you're creating a four new cons cells, whose cars are 1, 2, a, and b. The cdr of the final one is the existing list (3 4). That's because the implementation of append is something like this:
(defun append (l1 l2)
(if (null l1)
l2
(cons (first l1)
(append (rest l1)
l2))))
That's not exactly Common Lisp's append, because Common Lisp's append can take more than two arguments. It's close enough to demonstrate why all but the last list is copied, though. Now look at what that means in terms of your implementation of rev, though:
(defun rev (l)
(cond
((null l) '())
(T (append (rev (cdr l)) (list (car l))))))
This means that when you're reversing a list like (1 2 3 4), it's like you're:
(append '(4 3 2) '(1)) ; as a result of (1)
(append (append '(4 3) '(2)) '(1)) ; and so on... (2)
Now, in line (2), you're copying the list (4 3). In line one, you're copying the list (4 3 2) which includes a copy of (4 3). That is, you're copying a copy. That's a pretty wasteful use of memory.
A more common approach uses an accumulator variable and a helper function. (Note that I use endp, rest, first, and list* instead of null, cdr, car, and cons, since it makes it clearer that we're working with lists, not arbitrary cons-trees. They're pretty much the same (but there are a few differences).)
(defun rev-helper (list reversed)
"A helper function for reversing a list. Returns a new list
containing the elements of LIST in reverse order, followed by the
elements in REVERSED. (So, when REVERSED is the empty list, returns
exactly a reversed copy of LIST.)"
(if (endp list)
reversed
(rev-helper (rest list)
(list* (first list)
reversed))))
CL-USER> (rev-helper '(1 2 3) '(4 5))
(3 2 1 4 5)
CL-USER> (rev-helper '(1 2 3) '())
(3 2 1)
With this helper function, it's easy to define rev:
(defun rev (list)
"Returns a new list containing the elements of LIST in reverse
order."
(rev-helper list '()))
CL-USER> (rev '(1 2 3))
(3 2 1)
That said, rather than having an external helper function, it would probably be more common to use labels to define a local helper function:
(defun rev (list)
(labels ((rev-helper (list reversed)
#| ... |#))
(rev-helper list '())))
Or, since Common Lisp isn't guaranteed to optimize tail calls, a do loop is nice and clean here too:
(defun rev (list)
(do ((list list (rest list))
(reversed '() (list* (first list) reversed)))
((endp list) reversed)))
In ANSI Common Lisp, you can reverse a list using the reverse function (nondestructive: allocates a new list), or nreverse (rearranges the building blocks or data of the existing list to produce the reversed one).
> (reverse '(1 2 3))
(3 2 1)
Don't use nreverse on quoted list literals; it is undefined behavior and may behave in surprising ways, since it is de facto self-modifying code.
You've likely run out of stack space; this is the consequence of calling a recursive function, rev, outside of tail position. The approach to converting to a tail-recursive function involves introducing an accumulator, the variable result in the following:
(defun reving (list result)
(cond ((consp list) (reving (cdr list) (cons (car list) result)))
((null list) result)
(t (cons list result))))
You rev function then becomes:
(define rev (list) (reving list '()))
Examples:
* (reving '(1 2 3) '())
(3 2 1)
* (reving '(1 2 . 3) '())
(3 2 1)
* (reving '1 '())
(1)
If you can use the standard CL library functions like append, you should use reverse (as Kaz suggested).
Otherwise, if this is an exercise (h/w or not), you can try this:
(defun rev (l)
(labels ((r (todo)
(if todo
(multiple-value-bind (res-head res-tail) (r (cdr todo))
(if res-head
(setf (cdr res-tail) (list (car todo))
res-tail (cdr res-tail))
(setq res-head (list (car todo))
res-tail res-head))
(values res-head res-tail))
(values nil nil))))
(values (r l))))
PS. Your specific error is incomprehensible, please contact your vendor.
I am a LISP newbie.
To get the running sum of a list, I am writing like --
(setf sum 0.0)
(mapcar #'(lambda(x)
(setf sum (+ sum x)) sum) values))
For example, if you give '(1 2 3 4) as input, the above code returns '(1 3 6 10) as output and so forth.
Is it possible to do the same thing (in a more elegant way) without using the global variable sum ?
(loop for x in '(1 2 3 4) sum x into y collect y)
scanl is a oneliner:
(defun scanl (f init xs)
(loop for x in xs collect (setf init (funcall f init x))))
You could use loop, like this:
(defun running-sum (xs)
(loop with sum = 0
for x in xs
collect (setf sum (+ sum x))))
(running-sum '(1 2 3 4))
It's fundamentally the same thing, but it uses a local variable instead of a global one, and might be more clear.
Alternatively, you could define a recursive function, and a wrapper function:
(defun running-sum-recursive (xs)
(running-sum-recursive2 0 xs))
(defun running-sum-recursive2 (sum xs)
(if (eq xs nil)
nil
(let ((new-sum (+ sum (car xs))))
(cons new-sum (running-sum-recursive2 new-sum (cdr xs))))))
(running-sum-recursive '(1 2 3 4))
However this seems needlessly complicated to me when loop is available.
Note that in Haskell, you could do a running sum like this:
runningSum xs = scanl1 (+) xs
runningSum [1, 2, 3, 4]
The key here is the scanl1 function. It's possible that something similar exists in Lisp (and we've very nearly written it twice now), but I haven't used Lisp in a while.
Edit: After some searching, I don't think Common Lisp includes anything quite like scanl or scanl1, so here they are:
(defun scanl (f val xs)
(loop for x in xs
collect (setf val (funcall f val x))))
(defun scanl1 (f xs)
(cons (car xs)
(scanl f (car xs) (cdr xs))))
(scanl1 #'+ '(1 2 3 4))
Edit: Thanks to huaiyuan's answer for a suggestion about how the loops could be shortened.
Or you could use higher-order functions
(define (running-sum ls)
(cdr (reverse (foldl (lambda (y xs) (cons (+ (car xs) y) xs)) '(0) ls))))
Haskell does have a rich inventory of functions for list recursion, but we've got reduce at least. Here is an elementary (i. e. without the loop magic) functional solution:
(defun running-sum (lst)
(reverse (reduce (lambda (acc x)
(cons (+ (first acc) x) acc))
(rest lst)
:initial-value (list (first lst)))))
I'm using the head of the original list as the initial value and walk through the rest of the list adding sums at the head (because it's natural to add at the head), finally reversing the list thus obtained.
One can use reduce in most cases when there's a need to traverse a sequence accumulating a value.
Here is an elementary iterative solution using the push-nreverse idiom:
(defun running-sum (lst)
(let ((sums (list (first lst))))
(dolist (x (rest lst))
(push (+ x (first sums)) sums))
(nreverse sums)))
In Scheme I would calculate the sum of the list recursively using an accumulator. Like so:
; Computes a list of intermediary results of list summation
(define list-sum
(lambda (l)
(letrec ((recsum (lambda (lst acc acclst)
(if (pair? lst)
(recsum (cdr lst) (+ acc (car lst)) (cons acc acclst))
(cons acc acclst)))))
(recsum (cdr l) (car l) '()))))
Output:
> (list-sum '(1 2 3 4))
(10 6 3 1)
> (list-sum '(2 4 6 8 10))
(30 20 12 6 2)
>
The trick to recurse over a list is to take the first element/car off each time and pass the rest/cdr. You can keep intermediary results by using an extra parameter (called an accumulator) and pass the sum in that. I've used two accumulators above: one for the last sum and one for a list of all previous sums.
I've never done anything in LISP, so I can't tell if this translates directly to your dialect(?), but it's conceptually simple and I'm sure it's doable in LISP as well.
Do ask if something is not immediately clear. It's been a while since I've used this family of languages :)
I apologize for the bad English..
I have a task to write a function called "make-bag" that counts occurences of every value in a list
and returns a list of dotted pairs like this: '((value1 . num-occurences1) (value2 . num-occurences2) ...)
For example:
(make-bag '(d c a b b c a))
((d . 1) (c . 2) (a . 2) (b . 2))
(the list doesn't have to be sorted)
Our lecturer allows us to us functions MAPCAR and also FILTER (suppose it is implemented),
but we are not allowed to use REMOVE-DUPLICATES and COUNT-IF.
He also demands that we will use recursion.
Is there a way to count every value only once without removing duplicates?
And if there is a way, can it be done by recursion?
First of, I agree with Mr. Joswig - Stackoverflow isn't a place to ask for answers to homework. But, I will answer your question in a way that you may not be able to use it directly without some extra digging and being able to understand how hash-tables and lexical closures work. Which in it's turn will be a good exercise for your advancement.
Is there a way to count every value only once without removing duplicates? And if there is a way, can it be done by recursion?
Yes, it's straight forward with hash-tables, here are two examples:
;; no state stored
(defun make-bag (lst)
(let ((hs (make-hash-table)))
(labels ((%make-bag (lst)
(if lst
(multiple-value-bind (val exists)
(gethash (car lst) hs)
(if exists
(setf (gethash (car lst) hs) (1+ val))
(setf (gethash (car lst) hs) 1))
(%make-bag (cdr lst)))
hs)))
(%make-bag lst))))
Now, if you try evaluate this form twice, you will get the same answer each time:
(gethash 'a (make-bag '(a a a a b b b c c b a 1 2 2 1 3 3 4 5 55)))
> 5
> T
(gethash 'a (make-bag '(a a a a b b b c c b a 1 2 2 1 3 3 4 5 55)))
> 5
> T
And this is a second example:
;; state is stored....
(let ((hs (make-hash-table)))
(defun make-bag (lst)
(if lst
(multiple-value-bind (val exists)
(gethash (car lst) hs)
(if exists
(setf (gethash (car lst) hs) (1+ val))
(setf (gethash (car lst) hs) 1))
(make-bag (cdr lst)))
hs)))
Now, if you try to evaluate this form twice, you will get answer doubled the second time:
(gethash 'x (make-bag '(x x x y y x z z z z x)))
> 5
> T
(gethash 'x (make-bag '(x x x y y x z z z z x)))
> 10
> T
Why did the answer doubled?
How to convert contents of a hash table to an assoc list?
Also note that recursive functions usually "eat" lists, and sometimes have an accumulator that accumulates the results of each step, which is returned at the end. Without hash-tables and ability of using remove-duplicates/count-if, logic gets a bit convoluted since you are forced to use basic functions.
Well, here's the answer, but to make it a little bit more useful as a learning exercise, I'm going to leave some blanks, you'll have to fill.
Also note that using a hash table for this task would be more advantageous because the access time to an element stored in a hash table is fixed (and usually very small), while the access time to an element stored in a list has linear complexity, so would grow with longer lists.
(defun make-bag (list)
(let (result)
(labels ((%make-bag (list)
(when list
(let ((key (assoc (car <??>) <??>)))
(if key (incf (cdr key))
(setq <??>
(cons (cons (car <??>) 1) <??>)))
(%make-bag (cdr <??>))))))
(%make-bag list))
result))
There may be variations of this function, but they would be roughly based on the same principle.
I have the following filter function that filters out a list, x, that doesn't satisfy the function f.
For example, I call (filter 'evenp '(0 1 2 3)) and get back (NIL 1 NIL 3). But this is exactly my problem. How do I make it so that I just get back (1 3) ?
(defun filter (f x)
(setq h (mapcar #'(lambda (x1)
(funcall f x1))
x))
(mapcar #'(lambda (a b)
(cond ((null a) b)))
h x))
i.e. the problem is right here: (lambda (a b) (cond ( (null a) b) ) ) In my cond I don't have a t , or else statement, so shouldn't it just stop right there and not return nil ? How do I make it "return" nothing, not even nil, if the (cond ( (null a) b) ) isn't satisfied?
Much appreciated. :)
Based on this question it would be:
(remove-if #'evenp '(0 1 2 3))
Ignoring the other questions raised by this post, I'll say that mapcar will always return something for each thing it's mapping over, so you can't use another mapcar to clean up the NILs there. This is what mapcar does -- it walks over the item (or items, if mapping on multiple lists, as your second attempted mapcar does) and collects the result of calling some function on those arguments.
Instead, in this situation, if you had to use mapcar for some reason, and didn't want the NILs, you could use the remove function, i.e. (remove nil (mapcar ...))
Since #stark's answer is posted above, I'll say that the remove-if function there is essentially what you're trying to implement here. (That's where the question of whether or not this is for homework becomes most relevant.)
To answer the more general question of how to splice an arbitrary number of items (including none at all) into the result, mapcan (which is semantically mapcar + append) is useful for that:
(defun filter (f xs)
(mapcan (lambda (x)
(if (funcall f x)
(list x)
nil))
xs))
mapcan is also useful when you want to map an item to multiple results:
(defun multi-numbers (xs)
(mapcan (lambda (x) (list x (+ x x) (* x x))) xs))
(multi-numbers (list 1 2 3))
;=> (1 2 1 2 4 4 3 6 9)